This renewal application is focused upon two issues that are central to finding genes in the human genome. The first of these relates to the development of technology refinements to improve the selection and extension of cDNAa. Full length cDNAs are valuable tools in the positional cloning of disease genes, for annotating genomic DNA sequences and for functional genomics. However, no quick routes currently exist for moving from a cDNA fragment to a full length cDNA. We propose developing enhancements to increase the length of direct selected cDNAs and to enable the rapid conversion of cDNAs to full length. Direct cDNAs selection has been successfully applied to entire human chromosomes. We now propose applying these techniques to derive normalized cDNAs using the whole human genome. These types of cDNA libraries would complement and enhance existing normalized cDNA resources. The second issue that this application addresses is the development and mapping of new Expressed Sequence Tags (ESTs). The catalogue of human genes is far from complete, and gene prediction algorithms based upon genomic DNA sequences are not fully reliable. We propose to build a detailed map of 2,000 genes on human chromosomes 5 which is biased towards sequences that are not in the EST database. This will be facilitated by techniques we have developed that enrich for cDNAs which other methods miss. In our final aim we propose a limited scale up of this approach to derive new chromosome-specific cDNA libraries and novel EST maps for human chromosomes 3 and 8.